Systematic data-driven meta-analysis of class F fly-ash geopolymer concrete

• 799 fly-ash geopolymer mixes normalized to 150 × 300 mm cylinder strength • Dataset bundles chemistry, curing, multi-age strengths and CO₂ in one file • Optimal range: 10–12 M NaOH, silica modulus 1.5–2.3, curing ≤ 75°C • Design charts halve CO₂ versus OPC for ≥50 MPa structural concretes Geometry-dependent strength reporting has hindered reliable design of Class F fly-ash geopolymer concrete (FA-GPC). In this study, compressive-strength (CS) results from about 800 mixtures were normalized to the reference 150Ø300 mm cylinder, permitting direct cross-comparison of specimen geometries extracted from 67 peer-reviewed papers. The harmonized dataset bundles oxide chemistry, mix proportions, activator composition, curing schedules, fresh-state metrics, multi-age strengths, and cradle-to-gate CO₂ inventories. Normalized CS across 1–365 days spans 5.8–85.0 MPa, and CO₂ intensities range 66–895 kg CO₂/m³ (mean 160 ± 91). Data mining isolates practical activation windows—NaOH 10–12 M, silica modulus 1.5–2.3, curing ≤ 75°C—that consistently deliver 28-d CS ≥ 50 MPa at ∼160 kg CO₂/m³. Compared to strength-matched OPC concretes, these mixes reduce embodied carbon by ∼45–55% (median ≈ 50%). Strength–carbon design maps and the open dataset enable practitioners to target structural classes under explicit CO₂ budgets and provide a reproducible springboard for machine learning-based prediction, life-cycle assessment, and optimization of alkali-activated concretes and geopolymers.